Overview

ScaleOut Digital Twins™ provides a comprehensive user interface (UI) for deploying and managing digital twin models, connecting to data sources through popular event hubs or REST for streaming analytics, running digital twin simulations, and tracking aggregate statistics using graphical widgets. The UI is designed make it easy to take advantage of digital twins, while allowing ScaleOut Digital Twins to take care of all the details required for high performance and scalability.

When connecting to ScaleOut’s Azure-hosted cloud service, first create a cloud service account using the UI. Your account will initially be disabled until authorized by ScaleOut Software. Please contact ScaleOut Software (sales@scaleoutsoftware.com) to discuss billing options and authorize your account. You can start with an individual user account for evaluation purposes and later migrate to a multi-user, organizational account as described here.

Instead of using the cloud service, you can download the UI and connect it to an on-premises deployment of ScaleOut StreamServer®. Please see following topic for details on deploying and configuring the UI on-premises.

Note

Note that real-time digital twin models and connectors also can be deployed on-premises using APIs and ScaleOut StreamServer® instead of using the cloud UI. Please refer to the topic On-Premises Deployment for details.

After deploying the UI, you are ready to build and deploy a digital twin model following these steps:

  • Build a digital twin model in the language of your choice (Java, C#, or rules-based).

    You can build a Java or C# model with the ScaleOut Digital Twin Builder Software Toolkit, as described in this topic. You can also build a rules-engine or machine learning model using the ScaleOut Model development Tool, as described in the topic Model Development Tool.

  • Deploy a digital twin model using the UI as described here.

    The UI will ask you to upload the necessary binary files (usually in the form of a zip file) required to deploy the model. It will then send the files to the execution platform and prepare the model to receive messages from data sources. Note that the model’s message processor can write log messages, which appear in the UI.

  • If you are using real-time digital twins for streaming analytics, connect to data sources as described here.

    For initial testing, you may want to write an application that directly sends messages to your model. The next step is to connect to actual data sources using popular event hubs, such as Azure IoT Hub. You can use the UI to connect to one of the supported event hubs or REST and supply the necessary parameters.

  • If you are creating a simulation, configure and start the simulation as described here.

    You need to deploy at least one digital twin model with simulation digital twins to run in simulation. These twins can also interact with real-time digital twins to model data sources and test streaming analytics.

  • If you want to automatically persist digital twin state, you can connect to a persistence provider as described here.

    Connectors allow you to connect both to message hubs and persistence providers. The ScaleOut Digital Twins service supports SQL Server and SQLite for persistence, and it also provides full integration with Microsoft Azure Digital Twins.

  • Create graphical widgets to track aggregate statistics as described here.

    A widget lets you aggregate and visualize a property for all digital twin instances of a given model. You can create a widget in the UI by specifying the property to be aggregated, the aggregation operator (average, min, max, or count), an optional “group-by” property field to create multiple aggregation groups, and the chart type (bar, column, pie, or line). The widget will update approximately every five seconds with the latest statistics.

  • Run queries that display multiple properties for selected digital twin instances as described here.

    You can create tabular or geospatially mapped query results. Queries can run once or continuously.

  • Track the status of your deployment as described here.

    You can track the number of instances for each model, the message rate, memory usage, and simulation status. You also can select any instance and examine its properties to see how they are changing over time.

Once your application’s models and widgets or queries have been deployed, the ScaleOut Digital Twins service automatically performs a data-parallel computation every few seconds for each widget or continuous query to refresh its results. The execution engine transparently scales its throughput using a cluster of servers to ensure fast processing at all times.

The following sections provide a tour of the cloud service’s UI and explain its features in detail.